1. It is not assumed in the present paper that a high score on a self-report measure of depressive symptoms is equivalent to a diagnosis of depression. The two are not equated because numerous studies have shown that self-report measures tend to overestimate rates of depression compared to diagnostic surveys (e.g., Campbell & Cohn, 1991; Hodges, 1990; Oliver & Simmons, 1985; Roberts, Lewinsohn, & Seeley, 1991). This point is important to clarify since certain authors (e.g., Coyne, 1994; Fechner-Bates, Coyne, and Schwenk, 1994) have conveyed the impression that Vredenburget al. (1993) argued for the equivalence of self-reported distress and diagnosed depression. In actual fact, Vredenburget al. provided a clear statement of the need to distinguish depressive symptoms and diagnoses in student and nonstudent samples.
2. The Hammen (1980) study is often discussed as evidence for the instability of depressive symptoms in students. However, Hammen also identified many students who met the criteria for a depressive disorder following a diagnostic interview. The detection of this group of depressed students led Hammen to conclude that “the clinical significance of depression in this group is not negligible” (p. 127). Thus, diagnosable depression exists in some students and it is not mild and transient.
3. One explanation for the large decrease in scores is that Hammen's (1980) subjects were tested in an anonymous group setting at Time 1 with no expectation of follow-up, but Time 2 involved individual interviews. Differences in public versus private testing procedures lead to differences in self-presentational concerns that can influence the validity of findings of research on psychological distress (see Stanton, Burker, & Kershaw, 1991).